How COULD upgraded Ethernet cables make a positive difference? What's behind it?
- Thread starter caesar
- Start date
Caesar,
There are essentially three primary technical concerns when connecting a DAC at the end of a digital signal chain. Concern one, is that no data errors be introduced by the connecting interface. Concern two, is that jitter not be introduced by the interface. Data errors are uncommon over short, relatively low rate, interface links such as are found in home systems. Jitter can be introduced by the digital signal interface via impedance mismatch (which applies to wired and optical interfaces both), or via common-mode noise coupling over the interface. Concern three, is that common-mode noise can also directly couple to the analog output stages of a DAC, or even to the analog stages of a following pre-amp or power amp, through a DAC's wired digital signal interfaces. Optical digital signal interfaces are free of concern number three.
If different DAC digital interface cables sound different it is most likely due to their exhibiting differing jitter introduction profiles, and/or directly coupling common-mode noise in to the analog stages of any following equipment.
There are essentially three primary technical concerns when connecting a DAC at the end of a digital signal chain. Concern one, is that no data errors be introduced by the connecting interface. Concern two, is that jitter not be introduced by the interface. Data errors are uncommon over short, relatively low rate, interface links such as are found in home systems. Jitter can be introduced by the digital signal interface via impedance mismatch (which applies to wired and optical interfaces both), or via common-mode noise coupling over the interface. Concern three, is that common-mode noise can also directly couple to the analog output stages of a DAC, or even to the analog stages of a following pre-amp or power amp, through a DAC's wired digital signal interfaces. Optical digital signal interfaces are free of concern number three.
If different DAC digital interface cables sound different it is most likely due to their exhibiting differing jitter introduction profiles, and/or directly coupling common-mode noise in to the analog stages of any following equipment.
jkeny is correct. Noise that's harmful to a system does not have to be heard as noise to have an affect. In general if you have artifacts that climb with volume, something is broken. Low level fizzle or hum doesn't mean much of anything if you hear it with your ear at the speaker. The noise that's destructive is changing the music itself, happening at the transistors and other IC's. The result is the music sounds different, not full of noise. The very principle of a transistor explains why this is... one leg is signal, one is power, and one is output. The signal is what is telling transistor to shape the power as the music. If you have noise on the power it's not going to instruct itself, but when it's shaped the outcome is different. I hope those laymen terms are useful to understand the concept.
Impedance mismatch, AKA improperly made gear. Common-mode would be likely poor designed as well since you essentially need these conditions transformer PSU (small), poorly regulated voltage for DC offset between devices, poor rejection from smaller amount picked up by the cable, or poorly designed with safety grounds inappropriately connected to cause differences in potential. Good gear will have well regulated power, and reject common mode noise fairly well. Great gear will even have common mode chokes on the input.
I'm not certain that you get coupling from the ethernet to the analog; consider that what's noise for analog is carrier for it too. DSD has a carrier signal that's below ethernet's. A problem that could happen is the filtration meant to get rid of the DSD signal carrier may have inductors with SRF over DSD but under ethernet. Essentially you'd be sending all the ethernet signal right through. However this might be prevented also by a better receiver for ethernet since I believe there's a conversion from how DSD is sent over ethernet, to turn it into DSD. If the data is buffered up until then, the jitter over the ethernet is going to be entirely moot. Lastly the conversion device may have poor power, which could then couple to the analog section, or change it; but it's unlikely you're getting coupling from ethernet.
Then why is it being called noise?
OK, let's take radiated noise from the Ethernet cable. This interference is supposed to get into my high-end DAC even though I am assured by said DAC designer that the thousands of dollars if not tens of thousands of dollars I paid for that DAC over a 50 cent chip in my computer, is not subject to such vagaries. You are saying that this noise is not audible when the DAC is producing silence. But that when it plays music, it applies an amplification factor to that radiated noise? And that you now hear said noise in the midst of much louder music as opposed to silence? How could that be?
Can you tell me why I don't have recourse with my DAC designer if this is an issue? Is it an impossibility to design a DAC which doesn't have audible noise due to Ethernet cabling? What makes it impossible?
Do you want to explain what you mean by that question? It might help your understanding to expand the question?
It's called noise because it's not signal. There's more terms for types of noise. Noise is a term used in every field aside from audio, as they need to address it as well. It's not an audio specific term.
I'm sorry but who and when said radiated noise from the ethernet cable is getting into this DAC? Best as I can tell you're trying to twist concept and word. I explain how noise can change sound without being audible itself as an artifact. My example uses a transistor to try and make it simple. You turn that into radiation gets into your DAC? You've taken a step that's around 2,400 pages worth of textbooks past the example.
The ethernet cable itself will radiate, but it's not a big concern. It's going to be minimal since it's low voltage, short frequencies, and twisted wires. If it really could do much at all then you'd need shielded separate sections in the cable per twisted pair. Consider there isn't enough mutual sharing to cause problem except in very long runs.
If the DAC has a metal jack then with the enclosure it'll block pretty much everything besides the minimal amount that can exist radiating from the jack's pins as they travel to the board.
"You are saying that this noise is not audible when the DAC is producing silence. But that when it plays music, it applies an amplification factor to that radiated noise? And that you now hear said noise in the midst of much louder music as opposed to silence? How could that be?"
Producing silence? And actually I wasn't saying it applies an amplification factor, and I'm not sure why we'd be talking about radiated noise in that example to simply understand why noise changes sound but isn't audible as an artifact. First off noise that can be amplified will come from the signal. This is why fizzle with nothing playing doesn't change volume when it's from your amplifier, it's simply on the power that's pulling the electrons through the speaker to begin with (it's at max voltage already).
Typically unless something is broken high frequencies are essentially undetectable without playing anything (your speakers can't play 30mhz for example). But due to complex parasitic actions you might get some "radiation" affecting the device. At low signal level the more noise that's added the more it can be amplified, and the more it's amplified/buffered the more it will have parasitic potential. Because you can't hear the 30mhz, doesn't mean you can't hear the changes caused by it due to the interactions. In fact I hope you see the fallacy of thinking your speakers where going to play in excess of the frequency range any known speaker has ever been reported to be capable... oh, excuse me, "producing silence".
It's possible if a device has a very noisy power supply that it could change the signal pre-amplification due to "radiation" - again this type of noise would essentially constitute a broken device if it's spewing crap at high potential all over itself.
If a DAC has a poor design it's most likely to cause errors in the conversion process to DSD, than noise continue down the line. I dare not explain this, if I even can do a fair job since I'm not a designer of these chips. But consider at low signal level the parasitic qualities are much more destructive ("radiation"). And the exception being that the device may not filter the ethernet signal due to poor design (at this point it becomes noise, as signal is now a different mhz corresponding to DSD).
Overall I feel like I just wasted a lot of time on you being a troll, as I consider my original observation on how you chose to ask a question.
Last edited:
Oh, BTW...
I don't believe you were sincere about your dollars to performance statement/question either. We both know you find good and bad designs at all levels of price, and technical performance doesn't define subjective performance.
I don't believe you were sincere about your dollars to performance statement/question either. We both know you find good and bad designs at all levels of price, and technical performance doesn't define subjective performance.
I am just going by what Mike said:
What type do you think he meant? And why?
it was a suggestion by my son when I asked him what I could do to improve my Ethernet connection. he said that optical is much superior. I've not yet got it done yet. he has to order the pieces. he is our network engineer but I'm not in a position to judge his competence really.
enlighten me.....is this wrong?
Dodging your opportunity for self-learning, as usual, Amir
As I said you are blind to your own blindness & will always remain so once you adopt these types of self-defeating techniquesof avoiding learning.
You have had at least two opportunites to learn something in the last 2 pages but have avidly avoided both:
- You asked a specific question about noise which has a number of assumptions you need to state explicitly & you refuse to do so, this avoiding any attempt at learning
- you have presented measurements which are contradicted by Stereophile & you refuse to answer why.
To both of you - what power supply configurations are you using? I mean have you got PS isolators or conditioners into which equipment is plugged? Alternatively, are you using double insulated, class II equipment?
It might be useful to investigate the possible reasons for your immunity in this area?
I didn't hear Mike swap cables, nor pop the hood.
From a few things I do know, Lampizators are not faultless, and when in use once before, one version of the GG with volume was not a good match for my amp due to a poor output impedance. That problem isn't one with other equipment, or a buffer. Take that as you will, whether needing a buffer is ok or not on a flagship.
It sounded really good at Mike's. It isn't my favorite PCM DAC. And maybe other DAC's would have impressed me more, but without swapping DAC's it's hard to say more than it was really good. There is always a chance the associated equipment elevated it, but that's not so important, the overall experience matters and it was pretty awesome.
Steve Williams
Site Founder, Site Owner, Administrator
John
I certainly haven't tried as many as Bruce so my limited experience might taint your findings.
I'm interested in the questions posed in post 31 and the reply in post #32
Am I to infer from Amir's response that he is wrong and rather than saying that, he is saying "fine, let's move on"
I certainly haven't tried as many as Bruce so my limited experience might taint your findings.
I'm interested in the questions posed in post 31 and the reply in post #32
Am I to infer from Amir's response that he is wrong and rather than saying that, he is saying "fine, let's move on"
I would suppose that routers and switches, and perhaps their power supplies, have a much higher impact than cables. Did anyone try different brands and models?
Darn tooting. The man is masterful in what he does. Took just a few posts to break my back and cry uncle. I am at his mercy from here on:Am I to infer from Amir's response that he is wrong and rather than saying that, he is saying "fine, let's move on"
Glad someone caught on. The question for Steve and Bruce shouldn't be what cables they tried, but what equipment they use since they've nullified ethernet as a source of problems.
John
I certainly haven't tried as many as Bruce so my limited experience might taint your findings.
I'm interested in the questions posed in post 31 and the reply in post #32
Am I to infer from Amir's response that he is wrong and rather than saying that, he is saying "fine, let's move on"
One can judge this by the amount & continuous deflections he uses.
Last edited: